- /cold-email: Proven frameworks (AIDA, PAS, BAB) with 50-125 word limit
- /follow-up: Re-engagement psychology for follow-up sequences
- /readability: Flesch-Kincaid, Gunning Fog, SMOG metrics
- /word-stats: Word count, reading time, text statistics
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The `argument-hint` value containing `|` (pipe) was breaking YAML
parsing since `|` is special YAML syntax for multiline strings.
Adding quotes fixes the discovery issue.
Before: argument-hint: [text] [--intensity light|medium|aggressive]
After: argument-hint: "[text] [--intensity light|medium|aggressive]"
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Remove .claude-plugin/marketplace.json which was limiting skills.sh CLI
to only discover 1 skill. Without this file, the CLI does a recursive
search and finds both skills in skills/ folder.
Also simplified README installation instructions.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
According to Claude Code docs, skills in the skills/ directory are
automatically discovered when the plugin is installed. The 'skills'
field is for additional custom skill directories, not listing skills.
With source: "./", skills at ./skills/<name>/SKILL.md are auto-discovered.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- One plugin 'humanizerai' with source "./"
- Skills array lists ./skills/detect-ai and ./skills/humanize
- Each skill has SKILL.md directly in its folder
- This matches the working marketingskills structure
Install with:
/plugin marketplace add humanizerai/agent-skills
/plugin install humanizerai@humanizerai
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Claude Code plugins expect skills at:
<plugin>/skills/<skill-name>/SKILL.md
Restructured to:
- skills/detect-ai/skills/detect-ai/SKILL.md
- skills/humanize/skills/humanize/SKILL.md
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The old bundled skill was conflicting with separate detect-ai and humanize plugins.
Now only the individual skills remain:
- /detect-ai - AI detection
- /humanize - Text humanization
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The API returns score.overall (not score directly).
Updated documentation to show correct JSON structure
so Claude extracts the right score value.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Update plugin install commands to use separate detect-ai and humanize plugins
- Clarify that Settings is accessed via profile picture
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create skills/detect-ai/SKILL.md for /detect-ai command
- Create skills/humanize/SKILL.md for /humanize command
- Update marketplace.json to list both as separate plugins
- Update README with new installation commands
Install with:
/plugin marketplace add humanizerai/agent-skills
/plugin install detect-ai@humanizerai
/plugin install humanize@humanizerai
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove invalid 'skills' field from plugin entry
- Add plugin.json manifest for the humanizerai plugin
- Update installation instructions in README
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add .claude-plugin/marketplace.json for plugin marketplace integration
- Update README with improved installation instructions and documentation
- Add keywords to SKILL.md for better discoverability
- Bump version to 1.0.1
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Restructure to skills/humanizerai/ directory (name must match folder)
- Update SKILL.md frontmatter with license, metadata, allowed-tools
- Add LICENSE file (MIT)
- Update README with npx skills add installation option
- Add compatibility section for Claude Code, Cursor, Windsurf
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Romain SIMON